Integrated Application of the UASB Reactor and Ponds for Domestic Sewage Treatment in Tropical Regions

Waste stabilization ponds are widely applied for domestic sewage treatment in Brazil. The main objective of conventional waste stabilisation ponds (WSP's) is, nomen est omen , to remove the organic material from wastewater. To achieve this objective, a quite long liquid retention time ( R h ) is required: even under the favourable conditions of the tropics (high temperature, sunshine) a minimum liquid retention time of 20 to 30 days is necessary. It is well established that the optimal ponds configuration for domestic sewage treatment consists of a serial system composed of an anaerobic pond, followed by a partially aerobic (facultative) and one or more predominantly aerobic (maturation) ponds.By substituting the anaerobic pond by an efficient high rate anaerobic reactor like the Up Flow Anaerobic Sludge Bed (UASB) reactor, the pond receiving the effluent of this reactor - an polishing pond - will receive a relatively very low organic load and therefore the design criteria for this pond are very similar to those for maturation ponds, that is the polishing pond is designed for pathogen and/or nutrient removal.In semi-arid regions, like the North-East of Brazil, the use of treated sewage for irrigation water is an ecologically sound solution if precautions are taken to avoid contamination of the workers and of the consumers of the produced crops. According to the very sticky World Health Organization (WHO) standards for unrestricted irrigation, the helminth concentration must be less than 1 per litre and the faecal coliform ( FC ) concentration less than 10 3colony forming unit (CFU) per 100 ml. On the other hand, other important fact in design of pond systems is the evaporation from the pond needs to be minimal, and consequently the pond area must be as small as possible. When UASB effluent (with low organic material and suspended solids concentrations) is post-treated in polishing ponds, these systems can be designed specifically for FC removal. For that reason, their surface area can be kept smaller than that of conventional ponds, designed for organic material removal. For reducing the liquid retention time (and surface area) of a polishing pond it is necessary to minimize the degree of mixing as much as possible, so that the hydraulic regime approaches a plug flow regime.This thesis deals with aspects of design, performance and operation of sewage treatment systems consisting of an UASB reactor followed by a polishing pond. The thesis is directed to an experimental pilot plant investigation for assessing design and operation parameters with respect to the performance of the UASB reactor followed by post-treatment in different configurations of polishing ponds. With such a treatment system it was possible to produce an effluent that can be well used for unrestricted irrigation. The advantage of the system is that it allows a substantial reduction in the required surface area compared to the area required in conventional waste stabilization pond systems.As far as UASB reactor design concerned, it was shown that the phase separator design is of paramount importance to the performance of the system. The addition of parallel plates to the conventional phase separator could double the treatment capacity of the UASB reactor. The experimental investigation also showed that the sludge age and not the liquid retention time is the fundamental operational parameter of the UASB reactor: UASB reactors with different liquid retention times but the same sludge age will tend to have the same efficiency.An important aspect of UASB operation is the discharge of excess sludge. In this thesis a method was developed to estimate the frequency of excess sludge discharge as well as the optimal fraction of the total that can be discharged. It was shown that discharges of less than 50% of the sludge mass had a small and transitory effect on the UASB performance.Polishing ponds were operated both with a continuous and sequential batch feeding. Based on the experimental results of the investigations, relevant operational and conceptual aspects were elucidated and elements for dimensioning and designing polishing ponds could be provided.It was shown that the required liquid retention time for reduction of the BOD5 and TSS concentrations to a virtually constant value only amounts to about 3 days, which is very short i.e. a factor 7 to 10 times shorter than the liquid retention time (25 - 30 days) in conventional WSP's. In this investigation the FC concentration in the digested sewage was of the order of 10 7per 100 ml, so that the required removal efficiency for unrestricted irrigation ( FC < 10 3CFU /100 ml) was 99.99 % (4 log units). By using the value of the experimentally determined decay constant (2.0 to 2.2 day -1 at a depth of 0.65 m) the minimum time for this removal efficiency was calculated at 4 to 5 days for exponential decay. However the experimentally observed removal efficiency in a flow-through pond was much lower than expected on the basis of first order kinetics (exponential decay). This was attributed to the fact that partial mixing could not be avoided, even though the polishing pond was specifically designed to approach the plug flow regime as closely as possible. For this reason a minimum liquid retention time of about 10 days was required for a removal efficiency of 99.99 %.Mixing in continuous reactors like polishing ponds was quantified by determining the residence time distribution of the liquid by applying a slug of a tracer in the influent and observing its concentration as a function of time in the effluent. The experimental data obtained in these investigations showed that even in the pond that was carefully designed and operated to avoid mixing as much as possible, the dispersion number was of the order of 0.1 to 0.2, which means that mixing intensity was moderate. The tracer studies also revealed a surprisingly high dead volume fraction in the pond. When the values of the dispersion numbers and the dead volume fractions were used to calculate the theoretical faecal coliform removal efficiency in the pond from the Wehner and Wilhelm equation, it was established that there was a good correlation between theory and experimental data.Exponential decay for a first order process like FC removal can be guaranteed in a batch reactor. In this thesis the behaviour of ponds fed with sequential batches (SB) of anaerobically treated sewage was also evaluated. It was shown that the reduction in the FC concentration, was a good approximation, to the exponential decay of a first order process until a value of 1000 CFU /100 ml is attained, so that a very high removal efficiency (> 99.99%) of FC was obtained in a very short period of 4 to 5 days, as foreseen by theory for the experimental decay constant of 2.0 to 2.2 day -1. During this period complete elimination of helminth eggs also occurred. In addition, due to the efficient pre-treatment, an aerobic environment prevailed in the sequential batch (SB) ponds and a substantial further reduction of the BOD5 and TSS concentrations took place. Thus a final effluent was produced in ponds with a liquid retention time of only 4 to 5 days with a quality that complies with WHO standards for unrestricted irrigation. The liquid retention time required in a SB pond was only about half of the value for a flow-through polishing pond, so that its application leads to a reduction of the required pond area by a factor of 2. In conventional waste stabilisation pond systems the R h time under tropical climate conditions is of the order of 25 days, which is a factor 5 to 6 times greater than in polishing ponds operated in sequential batch mode for the production of an effluent for unrestricted irrigation.Other important finding of the investigations was that the rate of bottom sludge accumulation in polishing ponds remained much smaller than the values reported for conventional waste stabilization ponds, even though no sludge was discharged from the UASB reactor. The reduced bottom sludge accumulation has an important practical advantage: the frequency of desludging operation will be much lower than in anaerobic ponds and this operation may be unnecessary during the useful life span of the polishing pond.